LABORATORY CONNECTORS
Quality, Validation & Reconciliation
7 of 8Ensure laboratory data is accurate, consistent, complete and safe across Mediloop, HIS/EHR, LIS and instruments. Detect issues early, reconcile differences and maintain full traceability and auditability.
Data QualityConsistentSafeTraceableReconciledAuditable
Quality lifecycle overview
1
Message received
From LIS / instrument
2
Syntactic validation
HL7 / FHIR / schema
3
Semantic validation
Terminology & units
4
Business validation
Rules & consistency
5
Matching & correlation
Patient / order / specimen
6
Quality checks
Plausibility & completeness
7
Reconciliation
Compare & resolve
8
Acknowledgement
ACK / response
9
Stored & audited
Provenance & audit trail
Validation layers
Syntactic validation — message structure / schema
Semantic validation — codes, systems, units
Business validation — mandatory fields and logic
Value-set validation — LOINC, SNOMED CT, UCUM
Reference range and unit validation
Temporal validation
Patient / order / specimen consistency checks
Matching & correlation
Patient identity matching (MPI / demographics)
Order matching — Placer / Filler / Accession
Specimen matching — barcode / ID / attributes
Analyzer run / sample correlation
Test code mapping — local code ↔ LOINC
Container & collection time correlation
Duplicate detection & consolidation
Quality checks
Result plausibility — critical/high/low and delta checks
Reference range & unit compatibility
Duplicate / conflicting-result detection
Completeness — all expected results received
Outlier detection & sanity checking
QC flags from LIS / instrument
Specimen quality & rejection checks
Reconciliation
Compare LIS ↔ Mediloop ↔ HIS / EHR
Identify discrepancies & root cause
Auto-resolution with rules
Manual review for exceptions
Result correction / amendment handling
Status alignment — order, specimen, result
Reconciliation reports & dashboards
Acknowledgement & tracking
ACK / NAK for every received message
Order / result lifecycle status tracking
Pending, in progress, completed, cancelled
Retry with exponential backoff
Timeout detection & alerting
End-to-end delivery assurance
Exceptions & resolution
Exception queue for failed validations
Missing results / partial results
Unmatched orders or specimens
Duplicate or conflicting results
Manual resolution workflow
Escalation rules & notifications
Exception metrics & trending
Provenance & audit
Provenance for every result — who/what/when
Source system, device and operator
Original values & codes preserved
All transformations recorded
Full audit trail — create/update/correct
Immutable audit log ready
User access & data-access logging
Terminology & semantic layer — MSIS
LOINC for tests & observations
SNOMED CT for specimens & body sites
UCUM for units of measure
ICD for clinical context where relevant
Terminology mapping & normalization
Preserve original codes & values
Alignment with MSIS semantic services
Common data quality rules — examples
| Rule type | Example |
|---|---|
| Mandatory | Patient ID, test code, result value, unit, collection time |
| Code validation | LOINC exists, code system supported |
| Value validation | Numeric range and coded values valid |
| Logical validation | If status = final, result value must exist |
| Cross-field | Unit must be compatible with test |
| Temporal | Result time ≥ collection time |
| Referential | Order must exist before result |
Data flow with reconciliation
Mediloop HIS / EHR
LIS
Analyzer / Instrument
Workstation Agent
Quality, Validation & Reconciliation Engine
Runs across the exchange layer with configurable validation rules, real-time and batch processing, alerting, dashboards/KPIs and audit/compliance.
Metrics & dashboards
Orders with results (%)
Results with critical values
Turnaround time (TAT)
Exceptions count & trend
Duplicate / conflict rate
Missing results rate
Reconciliation success rate
Best practices
Validate early and fail safely
Preserve original identifiers, codes and values
Automate reconciliation where rules are deterministic
Escalate unresolved mismatches for controlled review
Track quality KPIs and recurring exception patterns
Keep mapping and validation rules versioned and auditable